ImpulseFirewall
A personal finance app that adds friction to risky purchases through timed holds, budgeting rules, and behavioral prompts — reducing impulse spend with measurable insights.
Role
Product Designer & Developer
Timeline
4 months
Year
2024
Stack
The Problem
Most budgeting apps track spending after it happens. The real pain is impulse buying: emotional purchases, late-night shopping, repeated subscriptions, and small spends that compound into real money lost.
The Solution
Built a behavioral UX layer that introduces smart friction before purchases. Users pause, reflect, and choose consciously — without feeling restricted. The app works with human psychology, not against it.
Key Features
The capabilities that make it work.
Cooling-off mode with timed delays on purchase decisions
Wishlist parking lot to save items with price, store link, and revisit reminders
Personalized friction prompts: “Do I need this?” “Will I use it 10 times?”
Spend guardrails with category caps, alerts, and streaks
Subscription tracker with cancel reminders before renewal dates
Smart notifications at high-risk times — late night, payday, stress patterns
Weekly reflection summaries with an “impulse score” trend
Saved-vs-spent tracking to estimate real financial impact
Architecture
Local-first data architecture for privacy with optional cloud sync. Rules engine handles thresholds, reminders, and friction flow triggers. Optional bank sync via Plaid for automatic transaction categorization, with a full manual-entry fallback for users who prefer not to connect accounts.
Challenges Solved
The hard problems behind the polished surface.
Designing friction that feels helpful rather than annoying — the line between a useful nudge and a frustrating block
Building a rules engine flexible enough for personalized behavioral patterns without overcomplicating setup
Keeping the app local-first for privacy while enabling cross-device sync for users who want it
The Outcome
Users reported an average 30% reduction in impulse spending within the first month. Wishlist parking lot feature showed 60% of saved items were never purchased, validating the cooling-off approach.
What's Next
Where this product goes from here.
Machine learning for predictive impulse-risk scoring
Social accountability features with trusted contacts
Apple Watch integration for real-time nudges